Universal Words relationship question-answering from UNL Ontology

Abstract

Universal Networking Language (UNL) represents natural language sentences as a semantic network of Universal Words (UWs). UNL Ontology provides the semantic background using relations for each UWs. However UWs are formal and all information provided by the UNL Ontology may not be understandable for human without the UNL specification. Especially we need to ensure the UW dictionary editors can understand the UWs information stored in UNL Ontology. To understand the UWs information, this research propose a question-answering system to provide the relationship information by explaining UWs relations with natural language. Using this system, the editors can easily understand the relationship information stored in UNL ontology.